What will the Energy Transition Actually Require?

Thoroughly addressing climate change depends on decoupling modern economies from the industrial way of life centered on the extraction and consumption of fossil fuels. It requires somehow running a modern, wealthy, and technology-driven civilization on natural earth flows — sun and wind — and weaning ourselves off of stocks of natural resources. To accomplish this, we have to both use less energy and fewer resources in total while reorganizing core industries like electricity generation, transportation, manufacturing, etc. to maximize the percentage of energy generated by renewable resources. However, overhauling the driving force behind civilization for the last two hundred or so years requires carefully considering the interplay between the qualities of industrialism that have benefited us, like technology, quality of life, and freedom, while minimizing the byproducts that threaten its existence, like environmental and social costs.

Defining what it means to have an “energy-efficient” economy

It’s standard knowledge that energy use is strongly correlated with wealth. Developed countries with high GDP per capita have been enjoying the productivity benefits of burning fossil fuels since the industrial revolution. As climate change ushers in a sense of urgency in reducing carbon emissions, we now have to focus analytical efforts on explaining the variance of outcomes within this group of wealthy countries; this means looking into the political, geographical, and economic variables that impact the energy efficiency of developed economies.

In isolating these variables, we can better understand the relationship between political economy and energy consumption and therefore nudge societies in a sustainable, energy-efficient direction. This means approximating the energy efficiency of an economy via metrics like a high-GDP-to-energy-consumed ratio, or high-GDP-to-Co2-emitted, and finding the predictor variables that increase this ratio.

Existing models studying the relationship between environmental degradation and economic development struggle to explain the complexity of industrialized energy use. My first exposure to a model of the relationship between GDP and the environment came from the environmental Kuznets curve, shown below.

The Kuznets curve predicts that as economies industrialize and increase their per capita income, they increase pollution — which makes sense, given the energy needs inherent in economic growth and high standards of living. However, once per capita income reaches a certain level, the model predicts that citizens then hold government and polluting corporations accountable by supporting regulations that limit carbon emissions and environmental degradation.

Despite its convincing narrative, the empirical evidence supporting this theory has been mixed. For usage of some specific natural resources, like forestry, or especially hazardous types of pollution (like air and water), there has been a track record of success. However, the evidence shows that for larger, aggregate metrics, like total energy consumed and Co2 emitted, countries have not reached a point where they are consuming less energy (and therefore emitting less carbon) per capita.

Yet MIT researcher and futurist Andrew MacAfee recently came out with an optimistic book titled More From Less claiming that wealthy countries have, in fact, reached this critical turning point. MacAfee argues that we’ve essentially reached “peak stuff” — or in other words, that advanced economies have decoupled economic growth from energy and material consumption. He argues this “dematerialization” phenomenon, which we are starting to see in developed nations, is not only a reduction in energy consumed per dollar of GDP, but less energy and fewer resources overall.

The ultimate measure of economic efficiency is wealth generated per unit of energy consumed, so if MacAfee is correct, we would expect to see a steady decrease in energy consumed per dollar of GDP. However, the results of a quick check using World Bank data doesn’t support this argument; in the plot below, which shows energy consumed per dollar of GDP from high-income countries between 1971 and 2015, it appears the world’s wealthiest economies are consuming a steadily increasing amount of energy.

Even when you separate out Switzerland, the darling of many political economy theorists, the trend is the same.

Intuitively, MacAfee’s argument makes sense. Yes, the iPhone has made all sorts of old electronic gadgets obsolete and condensed hundreds of CDs into one device. We’ve definitely relegated a lot of economic output and measures of wealth to the digital world, perhaps decreasing our material consumption. The key drivers of growth in the United States are tech and finance; office buildings to house people working in services require much less energy than industrial manufacturing. Digitization has made this possible for many reasons, but most of all, according to The Economist data is the new oil — and its predominantly in the hands of Western tech giants. Cultural trends also support this thesis. Wealthy countries have fewer children once child mortality rates decline.

But still, there are problems with this argument. Why hasn’t the energy intensity of GDP gone down in high-income countries? Does “dematerialization” take into account the total carbon footprint of goods consumed and produced in developed countries? And finally, the ever-present Catch-22 of wealthy societies: making our economy more energy-efficient would reduce energy use but lower the price of energy, and encourage us to consume more. Is this what has happened?

In developed economies, we may or may not have decoupled economic growth and GDP, but it remains to be seen whether this is a sign we’ve turned the corner, or if we’ve outsourced the energy-intensive industries to developing countries with more plentiful natural resources and less regulation. If this is the case, the total carbon footprint of our economies has not only not decreased, but it is increasing because we need to move these goods around the world.

Another key point is that technology, or even digital technology, is not inherently energy-efficient or good for the planet. Yes, we can use machine learning to predict renewable energy output, create a digitized grid, and fund companies like KoBold Metals to find accessible rare earth minerals for the battery revolution. However, technology has always increased the complexity and amount of energy in the economy. The more machines we have, the more energy we use — this has been true for millennia. The iPhones replaced lots of materials, but we now have loads of data centers to store the pictures on that iPhone and we only use each phone for about two years before it ends up in a landfill as e-waste. Technology can also prolong the fossil fuel era; one of data science’s first breakthrough industries was in oil and gas, as smart people found overlooked resources throughout the earth’s crust to exploit.

So we know that wealth and energy consumption move in lockstep. Wealth is responsible for, or a byproduct of, much of the material progress we’ve seen in developed and developing countries alike, enabling world-changing technologies in important industries like healthcare, transportation, etc. But what about the other signs of progress, those that are more related to governance and cultural variables, like political stability, violence, freedom of the press, and democracy? Can these ideological or cultural variables explain energy consumption?

You could make a convincing argument that political variables impact how well economies process energy, betting that countries with higher levels of openness, government accountability, political stability and rule of law develop norms and institutions that limit emissions and use energy more efficiently. If you contrast a democratic and capitalist country with market mechanisms to control the distribution of energy in the economy, you might bet that the efficiency of a capitalist economy increases the ratio of GDP to energy consumed. Perhaps rule of law and market societies encourage citizens to move goods and services more efficiently by encouraging people to invest in the future, accumulate capital, and think for the long term without fear that a corrupt government or war will destroy their wealth.

But when a classmate and I built a multi-linear model with a collection of political variables, the model explained very little variance in a country’s energy-consumed-per-unit-of-GDP ratio (github link forthcoming). This could be because a political factors don’t have much impact on this metric, or it could be because these relationships are so complex that a linear model isn’t robust enough to properly quantify this phenomenon, which makes sense when counting all the exceptions to simple rules of thumb we might try to impose on the diverse array of nations around the world. Oh, so you think countries with high GDP per capita inherently obtain political freedom? See UAE and Singapore as counterexamples. Economic development leads to open and freer societies? Not the case in China, which has become even more autocratic as wealth increases. Politically advanced countries avoid the resource curse because they don’t produce high amounts of natural resources? The US is a world leader in fossil fuel exports. Government oil income leads to political corruption and lower energy efficiency? Not in Norway. These exceptions will always be around because of the unique features of geography, history, culture, and economics.

In any socioeconomic model, we’d like to maximize two parameters: energy efficiency per GDP and GDP per capita. We want an energy-efficient society, one that can be reasonably expected to run on renewable energy one day, but one that has the baseline level of energy and wealth needed to enjoy iPhones, stimulate Internet use, encourage people to have fewer children, and plan for the future — some of the pivotal characteristics of dematerialization.

So what model can we use to design a wealthy yet energy-efficient economy? It’s hard when there is no country that has become wealthy via industrialization, advanced the political freedoms of its constituents and raised the standard of living, then decarbonized and improved the energy efficiency of its economy.

We need a model that will identify the characteristics that lead to free, wealthy societies with high energy efficiency and renewable energy usage, then learn to maximize these parameters to create a “perfect country”. It is tempting to think that if we load up a model with all sorts of demographic, geographic, economic, cultural, and environmental data, perhaps an advanced form of AI can figure out the combination that allows us to have our cake and eat it, too.

For example, perhaps this model would spit out a democratic country with a strong civil service and rule of law, lots of mountains for hydropower and a services-based economy — a political and environmental utopia. But it is unclear if the structure of the global economy will allow this country to be independent of a fossil-fueled global civilization. Do people in this country use products made in other countries? Are their service-based businesses profiting by orchestrating industrial activity in other developing nations? Given the intertwined nature of economics, we can’t stamp this utopia as a true utopia because it is still encouraging and benefiting from natural resource consumption in other countries.

It is difficult to model a society that doesn’t exist yet, and it is even more difficult to model a society when the desired parameters — renewable energy and high quality of life — are complex, unclear, and full of caveats. This energy transition is attempting to break a bond between energy consumption and freedom that has grown together in lockstep since the industrial revolution. Vaclav Smil, a scientist and policy analyst specializing in energy, has a less optimistic interpretation of material consumption than MacAfee. Referring to the energy transition, he posits:

“Could such a shift be accomplished without eventually converting to a no-growth economy and reducing the current global population? For individuals, this would mean a no less revolutionary delinking of social status from material consumption.”

Truly evolving past a civilization fueled by stocks of natural resources requires serious structural overall — one that changes the long-standing relationship between energy, wealth and freedom.